PDB40 Economic Analysis of Diabetes Treatment Goals Defined by Polish Diabetes Association: How Much Does Cost-Effective Treatment Cost?

A478

VALUE IN HEALTH 14 (2011) A233-A510

1 IMS Health, London, UK, 2IMS Health, Basel, Switzerland, 3Eli Lilly & Company, Prague, Czech Republic, 4LILLY UK, Hampshire, UK

OBJECTIVES: Glucagon-like peptide-1 (GLP-1) receptor agonists are indicated to improve glycemic control in adults with Type 2 diabetes mellitus. The maximum daily licensed dosages in the UK are 20␮g and 1.8mg for exenatide and liraglutide respectively. In addition to factors such as glycaemic control, cost is an important consideration when selecting treatments. The aim of this analysis was to describe the real-world daily usage and cost of exenatide BID and liraglutide in the UK setting. METHODS: Data and study period: UK records between October 2008 and March 2011 from the IMS Dynamic Prescription database. This database captures data from pharmacy records (45% national coverage) of actual prescriptions dispensed, linked to individual patients (anonymised). Inclusion criteria: patients have filled a prescription for a GLP-1 receptor agonist at least twice during the study period; all key prescription fields are complete. The weighted average daily usage was calculated for each agent using the total volume of product dispensed and the number of patients filling prescriptions per month. Drug costs (British National Formulary 61, 2011) were applied to estimate average daily cost (ADC). Key assumptions: patients are not stockpiling or disposing of drug; each prescription equals one pack; patients are filling their prescriptions at the same pharmacy. RESULTS: Data was available for a total number of unique patients of 19,200 and 12,690 for exenatide BID and liraglutide (data available from July 2009) respectively. The average daily usage during the investigated time period was estimated to be 20.49␮g for exenatide and 1.51mg for liraglutide, with an estimated ADC of £2.53 and £3.29 respectively. CONCLUSIONS: Based on the data described, GLP-1 receptor agonists are being dispensed in amounts within an acceptable range of the maximum daily licensed dosage. The ADC appears to be 30% higher for liraglutide with an estimated additional daily spend of £0.76. PDB36 ESTIMATING THE AVERAGE ANNUAL COST OF TREATMENT WITH INSULIN FOR PATIENTS WITH TYPE 2 DIABETES MELLITUS Burslem K1, Das Gupta R1, Hussein A2 1 Boehringer Ingelheim Ltd, Berkshire, Berkshire, Brack, UK, 2Boehringer Ingelheim Ltd, Berkshire, Bracknell, UK

OBJECTIVES: To estimate the average annual cost of treating patients with type 2 diabetes mellitus with insulin including: the cost of insulin, test strips for selfmonitoring of blood glucose levels, and additional healthcare professional (HCP) time spent with patients following insulin initiation. The secondary objective was to describe insulin prescribing patterns in the UK. METHODS: For insulin and test strip costs a retrospective analysis of 2009/10 UK patient-level data was undertaken using Cegedim Strategic Data. Costs were applied using the BNF and MIMS. To estimate HCP resource use, 100 HCPs were surveyed on the number of contacts with insulin patients in the 3 years prior to and the 3 years post insulin initiation. Costs were applied using PSSRU 2010. RESULTS: A projected 24.5 million insulin items were prescribed to 400,000 patients, generating an estimated average annual insulin cost of £393 per patient. Long-acting and biphasic insulins together accounted for more than 75% of the total volume and costs of insulin prescribed; intermediate acting insulins accounted for 6% and 4% of the volume and costs respectively. A projected 4.5 million packs of test strips were prescribed to 360,000 patients, generating an estimated average annual cost of test strips of £180 per patient. Contact time across all HCPs peaked in the year following insulin initiation. There was an absolute increase of 8 contacts per patient in the 3 years post insulin initiation, representing an additional cost of £103 per patient. CONCLUSIONS: Insulin initiation increases the cost of care not only because of the insulin costs, but because of the package of resources that insulin requires. The estimated cost of insulin, insulin pens, needles and test strips is £609 per patient. The analysis suggests divergence from the NICE Clinical Guidelines 87 recommendation that first-line insulin therapy should be intermediate NPH insulin. PDB37 INJECTION OF LONG-ACTING SOMATOSTATIN ANALOGS: A COST CONSEQUENCE ANALYSIS FOR THREE EUROPEAN COUNTRIES Marty R1, Roze S1, Kurth H2 1 HEVA, Lyon, France, 2IPSEN, Boulogne Billancourt, France

OBJECTIVES: Long-acting somatostatin analogs (SSA) with product-specific formulation and means of administration are injected periodically in acromegaly and neuroendocrine tumor (NET) patients. The ready-to-use device Somatuline Autogel/Depot® reduces drug administration time by 80%. Its prefilled syringe also avoids the risk of clogging reported for octreotide LAR. A simple decision-analytic model aimed at estimating cost savings due to these differences in administration was developed for the UK, France and Germany. METHODS: The decision tree simulated four scenarios for SSAs Somatuline Autogel/Depot® and Sandostatin LAR®, injected by either hospital- or community-based nurses. Injection success depended on clogging event occurrence. In the case of clogging, the first dose was assumed to be lost and a second injection performed. Administration costs were valued based on average hourly nurse wages in addition to country-specific retail drug costs. Several simulations were run depending on the baseline risk of clogging, administration time, and their respective relative reduction due to use of Somatuline Autogel/Depot®. RESULTS: Costs per successful injection were less for Somatuline Autogel/Depot®, ranging from EUR 13 to EUR 44, EUR 52 to EUR 150 and EUR 107 to EUR 127 respectively for France, Germany and the UK. As the prices for both long-acting SSAs were the same in France, cost savings came 100% from differences other than drug prices. For Germany and UK, the proportions of savings due to lower clogging and administration time was estimated around 32% and 20%, respectively. Based on low and high country-specific patient cohort size estima-

tions for acromegaly and NETs, these costs savings per patient could lead to overall annual savings up to one million euros for France, six million euros for Germany, and four million euros for the UK. CONCLUSIONS: Widespread usage of the new pre-filled Somatuline device for injection of SSA might lead to substantial savings for healthcare providers across Europe. PDB38 ECONOMIC EVALUATION OF RANIBIZUMAB IN THE TREATMENT OF VISUAL IMPAIRMENT DUE TO DIABETIC MACULAR EDEMA IN AUSTRIA Brennig C1, Schöllbauer V1, Walter E1, Gallagher M2, Knudsen MS3 1 Institute for Pharmaeconomic Research, Vienna, Austria, 2Novartis Pharma AG, Basel, Switzerland, 3IMS Health, London, London, UK

OBJECTIVES: Diabetic macular edema (DME) is an ophthalmological complication of diabetes that may lead to visual impairment and blindness if left untreated, and even despite treatment with the current standard of care, laser coagulation. Currently, an estimated 2% of diabetics suffer from DME with vision loss. The aim of the study was to evaluate the cost-effectiveness of ranibizumab versus laser coagulation in the treatment of visual impairment due to DME. METHODS: A costeffectiveness analysis was simulated using a Markov model adapted for Austria. The model is based on the PHIII-RESTORE trial. Outcome measures were ‘Vision Years’ and QALY. Costs are year 2010 values. Direct medical costs comprise all treatment costs due to diabetic macular edema. The cost of blindness was incorporated using data from an Austrian cost-of-illness-analysis. The model time horizon was lifetime. The analysis was performed from the perspective of the Austrian health care system according to the Austrian Guidelines for Health Economic Evaluations. RESULTS: The model assumes 7 injections of ranibizumab in the first year and 4 injections in the second year, as well as 2 treatments with laser coagulation in the first year and one treatment in the second year. Lifetime costs amount to €17,417 for ranibizumab and to €16,286 for laser coagulation. The ICER is €5354 (incremental QALYs gain with ranibizumab of 0.22). The number of vision years is 10.19 for ranibizumab and 8.57 for coagulation; the incremental cost per additional vision year gained is €701. CONCLUSIONS: The study suggests that in Austria, ranibizumab treatment for visual impairment resulting from DME is a cost-effective strategy versus the current standard of care, laser coagulation. PDB39 COST-EFFECTIVENESS OF SAXAGLIPTIN COMPARED TO SITAGLIPTIN FOR THE TREATMENT OF PATIENTS WITH TYPE 2 DIABETES MELLITUS (T2DM) Hutchings A1, Tolley K2, Achana F3, Brereton NJ4, Lebmeier M5, Mesa OA6 1 GMAS, London, UK, 2Tolley Health Economics, Buxton, UK, 3University of Leicester, Leicester, UK, 4BresMed Health Solutions, Sheffield, South Yorkshire, UK, 5Bristol-Myers Squibb, Uxbridge, Middlesex, UK, 6Bristol-Myers Squibb Pharmaceuticals Ltd, Uxbridge, Middlesex, UK

OBJECTIVES: Saxagliptin (Onglyza®) and sitagliptin (Januvia®) are DPP-4 inhibitors licensed for the treatment of T2DM. The two treatments have been investigated as an add-on to metformin in an 18-week, non-inferiority, RCT in 801 patients with T2DM who failed to achieve adequate glycaemic control on metformin alone. Results showed that the newer treatment, saxagliptin, was noninferior to sitagliptin, with a similar tolerability profile. Saxagliptin has a lower acquisition price, hence this analysis sought to assess cost effectiveness of saxagliptin⫹metformin versus sitagliptin⫹metformin using a cost utility analysis (CUA) framework from a UK healthcare perspective. METHODS: The CUA utilised a validated model using UKPDS risk equations to estimate long run micro/macro-vascular complications and mortality over a 40 year time horizon. Clinical parameters in the model included HbA1c levels for treatment effect, weight gain and incidence of hypoglycaemic adverse events. Parameter estimates were obtained from a mixed treatment comparison (MTC) of saxagliptin and sitagliptin, which included the head-to-head study. Treatment costs were based upon UK published list prices. Established costs and disutilities associated with long-term diabetic outcomes were used, based upon a UKPDS sub study. Univariate/probabalistic sensitivity analysis was conducted. RESULTS: The annual drug cost per patient for saxagliptin was £411.93 versus £433.57 for sitagliptin. In the base case, total discounted healthcare costs over the 40 year time horizon were £9,907 with saxagliptin and £10,035 with sitagliptin, with the same discounted QALY outcomes (10.49). Saxagliptin was therefore cost saving in the base case analysis. This finding was consistent across a range of sensitivity analyses, with the exception of lower 95% credible intervals for saxagliptin efficacy which resulted in a small incremental cost for saxagliptin (£29). CONCLUSIONS: Saxagliptin and sitagliptin have been shown to have comparable therapeutic profiles in a head-to-head study and MTC, but lower healthcare costs driven by a 5% lower drug acquisition cost. PDB40 ECONOMIC ANALYSIS OF DIABETES TREATMENT GOALS DEFINED BY POLISH DIABETES ASSOCIATION: HOW MUCH DOES COST-EFFECTIVE TREATMENT COST? Szmurlo D1, Schubert A2, Kostrzewska K1, Rys P1, Skrzekowska-Baran I2 1 HTA Consulting, Krakow, Poland, 2Novo Nordisk Pharma Sp z.o.o., Warsaw, Poland

OBJECTIVES: Clinical guidelines for diabetes management issued by Polish Diabetes Association (PDA) describe therapeutical goals in patients with diabetes. The aim of this analysis was to determine additional costs that may be incurred for treatment along with PDA recommendations (as compared with current treatment practice), so that the growth of treatment-related expenses would remain costeffective in Polish setting. METHODS: Two hypothetical patients were defined: John and Peter, whose clinical characteristics correspond to those of newly diagnosed patients with diabetes mellitus type 2 (DM2) in Poland. Diabetes progression was modelled assuming that John is treated in line with current clinical practice and Peter is treated along with PDA recommendations (HbA1c, LDL, HDL, SBP are

VALUE IN HEALTH 14 (2011) A233-A510

maintained within PDA-defined limits). Simulations were conducted in CORE diabetes model, which is a Markov model built on the base of published clinical trials and encompasses over a dozen of diabetes complications. The model was extensively validated and allows for reliable estimation of costs and outcomes associated with diabetes. Model inputs were adapted to Polish setting. Economic analysis was conducted in lifetime horizon, costs and outcomes were discounted (5% and 3,5%, respectively). Cost acceptability threshold in Poland is 25 511 euro per QALY gained. RESULTS: John’s QALY is 0,3 lower that QALY of Peter. Treatment of John’s complications is 400 euro more expensive as compared to Peter. If willingness to pay (WTP) equals to €7500 euro per QALY, yearly costs of Peter’s treatment may be 250 euro higher that John’s. If WTP is €15,000, Peter’s treatment may be €450 more expensive that John’s and if WTP is €25,000 the difference in treatment costs may be as high as 725 euro. CONCLUSIONS: DM2 treatment along with PDA recommendations may be cost-effective provided additional costs do not exceed €725 per year. PDB41 THE COST-EFFECTIVENESS OF GETTING TO GLUCOSE, BLOOD PRESSURE, AND LIPID GOALS IN PATIENTS NEWLY DIAGNOSED WITH TYPE 2 DIABETES MELLITUS (T2DM) AND YOUNGER THAN FIFTY IN SWEDEN He J1, Neslusan C1, Willis M2, Worbes-Cerezo M3 1 Janssen Global Services LLC, Raritan, NJ, USA, 2IHE, Lund, Sweden, 3Janssen-Cilag S.A., Madrid, Spain

INTRODUCTION: Good T2DM management requires not only good control of blood glucose, but also blood pressure and serum lipid levels. Although data from the Swedish National Diabetes Registry indicates that more patients have attained recommended levels of these biomarkers over time, a sizable proportion fails to meet all of these goals. OBJECTIVES: Assess the cost-effectiveness of intensifying therapy to achieve Swedish-specific treatment goals for HbA1c, systolic blood pressure (SBP), and LDL versus usual care for patients newly diagnosed with T2DM and younger than fifty. METHODS: We used the Economic and Health Outcomes (ECHO)-T2DM model, a Markov-based micro-simulation model, to simulate the lifetimes of 500 cohorts of 500 hypothetical patients under two different scenarios: 1) treatment to maintain target goals for HbA1c, SBP and LDL; and 2) treatment to maintain levels observed empirically in Sweden. Pharmacotherapy treatment pathways for the control of hyperglycemia, hypertension and dyslipidemia followed Swedish guidelines and were identical in the two scenarios. The costs of pharmacotherapy and medical events were obtained from Swedish data. RESULTS: Treatment to HbA1c, SBP and LDL goals versus treatment to observed levels in Sweden resulted in a small QALY gain (0.13) and medical cost-savings of SEK 3552(€395). Spending on glucose-lowering agents, anti-hypertensives, and lipidlowering agents was increased by SEK 4136(€460), SEK 4864(€540) and SEK 2390(€265), respectively. Costs due to micro- and macrovascular complications were reduced by SEK 5731(€637) and SEK 9522(€1058), respectively. CONCLUSIONS: For patients newly diagnosed with T2DM and younger than fifty in Sweden, intensifying therapy to maintain target glucose, blood pressure, and lipid levels resulted in increased spending on pharmacotherapy, however, spending on micro- and macrovascular events was reduced by a greater degree. These results suggest that allocating more resources toward the attainment of these goals may be welfareimproving. PDB42 ECONOMIC EVALUATION OF RECOMBINANT HUMAN FSH IN COMPARISON WITH URINARY HMG IN ASSISTED REPRODUCTION IN THE GREEK SETTING Fragoulakis V1, Tarlatzis B2, Mastrominas M3, Maniadakis N1 1 National School of Public Health, Athens, Greece, 2Medical School, Aristotle University of Thessaloniki, Thessaloniki, Greece, 3Embryogenesis Centre for Reproductive and Fertility Studies, Athens, Greece

OBJECTIVES: To compare the cost-effectiveness of Follitropin Alpha (Gonal-F®), which is a recombinant FSH, with a urinary highly purified hp-FSH (Menopur®) used in assisted reproduction in Greece. METHODS: A decision tree in combination with a Markov model was constructed to assess the clinical and economical impact of comparators for three consecutives cycles. Transition probabilities for all stages of a treatment cycle (i.e, cancelled ovum retrieval, successful recovery of oocytes etc) were derived from literature and validated by clinical experts. Cost components such as “initial treatment cost”, cost of “oocytes”, “oocyte pick-up”, “fertilization”, “transfer”, “cryo preservation” and “frozen- thawed embryo transfer (FET)” were derived from the electronic databases of selected private and public clinics. The average number of units used per IVF and the rate of adverse events were based on the literature. Drug prices and reimbursement tariffs, were obtained from the “Government Gazette” and valued at 2011 prices. A probabilistic sensitivity analysis was performed to deal with uncertainty and to construct variability measures. RESULTS: There was a statistically significant difference in favor of the r-FSH arm compared to hp-HMG, which is associated with 52 more life births (95%CI: 26-78, p-value⬍0.001) per 1,000 patients. The cost per life birth was estimated at €16,906 (95%CI: €16,347 – €17,516) and €17,286 (95%CI: €16,740 – €17,845) in the r-FSH and hp-HMG arms, respectively. The cost per IVF was estimated at €4,365 (95%CI: €4,205 – €4,506) in the r-FSH and €3,815 (95%CI: €3,661 – €3,953) in hp-HMG arm, indicating a difference at €550 (95%CI: €365 – €730, p-value⬍0.001). The incremental cost per life birth (ICER) for r-FSH versus hp-HMG was estimated at €14,540 (95%CI: €10,509 – €21,868), while the incremental cost per life year was estimated at €4,153 (95%CI: €2,038 – €6,233). CONCLUSIONS: r-FSH may represent a cost-effective choice compared with a urinary hp-FSH (Menopur®) used for ovarian stimulation in the Greek setting.

A479

PDB43 THE ECONOMIC IMPACT OF WEIGHT LOSS IN PATIENTS NEWLY DIAGNOSED WITH TYPE 2 DIABETES MELLITUS (T2DM) AND YOUNGER THAN FIFTY IN SWEDEN Willis M1, Neslusan C2, He J2, Worbes-Cerezo M3 1 IHE, Lund, Sweden, 2Janssen Global Services LLC, raritan, NJ, USA, 3Janssen-Cilag S.A., Madrid, Spain

OBJECTIVES: This study estimated the effect of weight reduction on long-term outcomes and associated direct medical costs for patients newly diagnosed with T2DM and less than fifty years old in Sweden. METHODS: We simulated the lifetimes of 500 cohorts of 1000 patients with characteristics based on the Swedish National Diabetes Register using the Economic and Health Outcomes (ECHO)-T2DM model. All patients were assumed to increase weight over time (0.23 kg per year) however, half of the patients were assumed to lose 5 kg in the first year, so that a 5 kg differential was maintained. The effect of weight on T2DM complications was modeled using risk equations from the UK Prospective Diabetes Study, wherein weight is only a direct determinant of the risk of congestive heart failure (CHF). The risks of stroke and myocardial infarction are affected only indirectly via their linkage with CHF, and mortality risk is affected only indirectly via macrovascular event history. Weight change was assumed to impact QALYs by an amount reported in the T2DM-specific CODE-2 study. Pharmacotherapy was administered according to Swedish recommendations and Swedish cost data was used for medical events and pharmacotherapy. RESULTS: A weight loss of 5 kg resulted in cost-savings of SEK 654 (€69) over an average of 17.1 years, mainly attributable to reductions in CHF incidence. Life years increased marginally; QALYs, however, increased more substantially (0.18). CONCLUSIONS: At a relatively conservative willingness-to-pay threshold of SEK 250,000 (€26,540), an intervention that resulted in a one-time weight loss of 5 kg would be welfare improving at a cost of up to SEK 45,654 (€4,846) over 17.1 years. As this simulation conservatively excluded a number of other benefits of weight loss (e.g., effects via improved lipids, blood pressure and reductions in other weight-related illnesses), the true economic value is likely greater. PDB44 AN ECONOMIC EVALUATION OF THE USE OF PIOGLITAZONE IN ITALY USING PROACTIVE Gladwell D1, Capri S2, Fionda A3, Massi-Beneddeti M4 1 Takeda Global Research and Development Europe, London, UK, 2Institute of Economics, Castellanza, Italy, 3Takeda Italia Farmaceutici S.P.A, Rome, Italy, 4University of Perugia, Perugia, Italy

OBJECTIVES: The aim of this economic evaluation was to test the hypothesis that the clinical benefits observed with pioglitazone in the PROactive Study will lead to economic benefits in terms of reduced macrovascular complications costs and insulin treatment in Italy (the trial compared standard of care ⫹ pioglitazone versus standard of care alone). METHODS: Two analyses were undertaken; within trial analysis and life-time simulation. The PROactive study provided the clinical and resource utilization data to estimate the cost-effectiveness of pioglitazone in the within trial analysis and was the basis for the secondary analysis which undertook a life time simulation using a modified version of the validated CORE diabetes model. CODE-II utility values were used for the base case. Due to the distribution system of pioglitazone in Italy, two different prices were used; the public price paid by the retail market (€2.11 per patient per day) and the ex-factory price discounted by 25% (€ 0.96 per patient per day). Costs and health gains were discounted at the joint rate of 3%. RESULTS: The incremental utility gain in within trial analyses was 0.0191, the incremental event and medication costs in the public price scenario were €842 leading to an ICER of €43,996 per QALY. In the lifetime simulation model the incremental utility gain was 0.149, the incremental event and medication costs in the public price scenario were €3,783 leading to an ICER of €25,426 per QALY. In the ex-factory price discounted by 25% scenario the medication costs were lower leading to the inclusion of pioglitazone in treatment being dominant in both analyses. CONCLUSIONS: In the Italian setting reduced costs for macrovascular complications and insulin treatment leads to the inclusion of pioglitazone in treatment being within standard cost-utility thresholds and is therefore an effective use of health resources. PDB45 COST-EFFECTIVENESS OF TRANSFERRING TYPE 2 DIABETIC PATIENTS FROM NEUTRAL PROTAMINE HAGEDORN (NPH) TO DETEMIR IN PORTUGAL SETTINGS Carvalho D1, Lindner L2, Kozarzewski M3 1 São João Hospital, Porto, Porto, Portugal, 2IMS Health, Barcelona, Cataluña, Spain, 3Novo Nordisk Portugal, Lisboa, Lisboa, Portugal

OBJECTIVES: To estimate the long-term cost-effectiveness of transferring type 2 diabetes patients to an insulin detemir regimen therapy from a Neutral Protamine Hagedorn (NPH) insulin regimen in the Portuguese routine clinical practice. METHODS: A computer simulation model “CORE Diabetes Model” was used to make long-term projections of clinical outcomes and direct medical costs based on short term findings from the European cohort in the PREDICTIVE trial. Therapy conversion to insulin detemir was associated with a reduction in glycosylated haemoglobin (HbA1c) by 0.2% (p ⬍ 0.05), mean body weight was reduced by 0.7 kg (p⬍0.01) and the incidence of total hypoglycaemia decreased from 11.7 to 3.0 episodes per patient/year (p ⬍ 0.0001). Events were projected for a time horizon of 30 years. The cost analysis takes the perspective of the Portuguese National Health System. RESULTS: Therapy conversion to insulin detemir plus OADs improves life expectancy by 0.056 years and quality-adjusted life years (QALY) by 0.462 compared to NPH insulin plus OAD. The incremental cost effectiveness ratio cost per life years gained and per QALY gained with insulin detemir plus OADs treatment as compared to NPH insulin plus OADs is 3,239€ and 393€ respectively. Type 2 diabetes